The TRICLOBS Dynamic Multi-Band Image Data Set for the Development and Evaluation of Image Fusion Methods

Area of Science:

• Computer Vision
• Remote Sensing
• Sensor Fusion

Background:

• Multiband nighttime imagery fusion and enhancement are critical for surveillance and navigation.
• Existing datasets lack dynamic multiband imagery, hindering algorithm development.
• There is a need for comprehensive datasets that capture real-world surveillance scenarios.

Purpose of the Study:

• To introduce the TRICLOBS dynamic multiband image dataset.
• To provide a resource for developing and evaluating image fusion, enhancement, and color mapping algorithms.
• To support advancements in short-range surveillance applications.

Main Methods:

• Collected sixteen registered visual, near-infrared (NIR), and long-wave infrared (LWIR) motion sequences using the TRICLOBS system.
• Captured diverse military and civilian surveillance scenarios across three distinct scenes.
• Included static and dynamic elements such as people, vehicles, foliage, and buildings.

Main Results:

• The TRICLOBS dataset offers dynamic, registered multiband imagery (Visual, NIR, LWIR).
• It encompasses various surveillance scenarios with diverse targets and backgrounds.
• The dataset includes color photographs for realistic color remapping development.

Conclusions:

• The TRICLOBS dataset fills a critical gap for dynamic multiband image analysis.
• It enables the development and evaluation of advanced fusion, enhancement, and color mapping algorithms.
• Facilitates the creation of more realistic and effective surveillance systems.